Withdrawal Averaging?

[marko]

Different kind of question (I think).

My 10 year W/D average has been about 2.7% including YTD.

My calculated SWR is about 4%.

Due to a variety of reasons, the past two year's withdrawals have been in the 6% range. We just got hit with a lot of 'stuff' in '17 and '18. (all good stuff BTW)
I expect that next year and following years will drift back down to 4% or lower.

I think that's ok as we've essentially 'banked' 1.3% per year over the previous 10 years. The previous 8 years averaged a W/R of 1.7%.

Is my thinking out in the weeds or does this make sense? Can you/should you average your withdrawals over time?

 

[audreyh1]

Are you calculating your annual withdrawal based on your initial portfolio value and adjusting for inflation each year?

Or are you calculating what 4% of your portfolio is every year and using that?

If it’s the latter, where your income drops if you portfolio drops, I wouldn’t worry about it and just take ~4% of whatever is remaining.

 

[RunningBum]

As long as they are actually exception expenses, and not all lifestyle creep. Banking reserves in low spending years, whether actually setting it aside, or just doing it on paper, and using those reserves when you have extraordinary expenses, makes perfect sense. You probably still have a decent amount left in those reserves.

 

[marko]

Are you calculating your annual withdrawal based on your initial portfolio value and adjusting for inflation each year?

Or are you calculating what 4% of your portfolio is every year and using that?

If it’s the latter, where your income drops if you portfolio drops, I wouldn’t worry about it and just take ~4% of whatever is remaining.

I'm looking at what I withdrew each year compared to the starting balance of that year. I know that's not the basis of SWR concept, now that you mention it. But the portfolio growth has been positive YoY over the past decade despite withdrawals.

I'm sort of thinking that with a 1.7% WR over the past 8 years, it's fairly negligible and perhaps my 'real' start/withdrawal balance began two years ago when I started withdrawing serious percentages.

 

[marko]

As long as they are actually exception expenses, and not all lifestyle creep. Banking reserves in low spending years, whether actually setting it aside, or just doing it on paper, and using those reserves when you have extraordinary expenses, makes perfect sense. You probably still have a decent amount left in those reserves.

That's kind of what I'm thinking. A bit of lifestyle creep as we got more comfortable with retirement and the market.

But I envision us creeping up to our 4% SWR from 1.7%; not staying at 6%.

Part of that 6% was paying a portion of a nephew's final two years of college, two new cars, and a kitchen reno...all stuff that won't recur going forward.

 

[Dtail]

So you are using the remaining portfolio % methodology effectively.
Why not model it in Firecalc in the methodology section (instead of the constant spending methodology) and fill in all the regular sections including your expected remaining years?

 

[RunningBum]

That's kind of what I'm thinking. A bit of lifestyle creep as we got more comfortable with retirement and the market.

But I envision us creeping up to our 4% SWR from 1.7%; not staying at 6%.

Part of that 6% was paying a portion of a nephew's final two years of college, two new cars, and a kitchen reno...all stuff that won't recur going forward.

College, certainly not, unless you have more relatives to help.

Cars will probably be replaced again some day, so you'd be back to building up a reserve for those again.

You're probably done with your kitchen, but your house may need other work, so that's another reserve to fund.

I'm just saying if you drop back to 4% with no special expenses, you're really above 4% because you should be funding, at least on paper, a reserve for those things.

Then again, between underspending the first 8 years, good market returns, and being 10 years older, you can probably spend more than 4%. Just beware that the market has been roaring, and eventually a drop is probably coming. Strike that, one certainly is coming, but who knows when, how big, and for how long.

 

[marko]

So you are using the remaining portfolio % methodology effectively.
Why not model it in Firecalc in the methodology section (instead of the constant spending methodology) and fill in all the regular sections including your expected remaining years?

I would but I view the past two years as a blip more than a regular thing. My SWR per FireCalc and others is ~4%.

I was more interested if you essentially 'bank' the amount of the 4% that you don't use each year.

Except for the past 2 years (out of my 14 years RE'd) we've barely touched the portfolio (~1.7% per year) and some years not at all; I don't really consider us having started our SWR.

So maybe what I do need to do is figure out WHEN to hit start for a true read on a 'SWR plus inflation' plan.

 

[TwoByFour]

That's kind of what I'm thinking. A bit of lifestyle creep as we got more comfortable with retirement and the market.

But I envision us creeping up to our 4% SWR from 1.7%; not staying at 6%.

Part of that 6% was paying a portion of a nephew's final two years of college, two new cars, and a kitchen reno...all stuff that won't recur going forward.

This was the exact point I made in another thread "Withdrawal variability" - that what is important is the average because a lot of expenses that go into the long term average occur very infrequently. A 6% withdrawal last year is fine as long as the average is less than your target.

BTW, a note on terminology. I always think of SWR as the maximum one can withdrawal, which is not necessarily your actual, target rate of withdrawal. We picked a WR based on a budget and that is our target. It is 3.3%. Most years it is less, some it is more. I have no idea what our SWR is. Based purely on Trinity, it would be 4%, but I take that study with a grain of salt since it is a bit dated.

 

[marko]

College, certainly not, unless you have more relatives to help.

Cars will probably be replaced again some day, so you'd be back to building up a reserve for those again.

You're probably done with your kitchen, but your house may need other work, so that's another reserve to fund.

I'm just saying if you drop back to 4% with no special expenses, you're really above 4% because you should be funding, at least on paper, a reserve for those things.

Then again, between underspending the first 8 years, good market returns, and being 10 years older, you can probably spend more than 4%. Just beware that the market has been roaring, and eventually a drop is probably coming. Strike that, one certainly is coming, but who knows when, how big, and for how long.

Agreed 100%. But our spending/income profile will change for the better in the next few years (DW going to Medicare and taking SS--$35K right there), some other fixed expenses ending and a hefty inheritance in the wings. (yes, the inheritance is locked in)

 

[Red Badger]

We pulled ~12% last year. From that, we built a CD ladder to get us to late 2019 (when we'll both file for SSI). Then, we'll be happy with 2.5 to 3% WR (which is mostly our discretionary bucket).

In a severe downturn, we'd probably reduce/rebalance our WR in the hopes of having a larger base to grow upon a market recovery.

I'm in the camp that agrees that knowing your YOY WR average is helpful if one has lumpy years. We anticipate some lumpy years ahead, as we will plan some big ticket travel / experiences that will be impractical in later years . That's assuming we're around for those later years.

 

[joeea]

Is my thinking out in the weeds or does this make sense? Can you/should you average your withdrawals over time?

Whatever you have done in the past makes absolutely no difference going forward.

Examine your assets. Determine what rate of withdrawal you would like. Analyze to see if your assets can support that withdrawal rate going forward. If not, adjust accordingly. Repeat the process periodically.

Average doesn't matter. People drown in rivers that average 2" in depth.

 

[MichaelB]

You may be overthinking this. Even if it there is a bit of lifestyle creep, that's not necessarily a problem. After all, the total burden your portfolio has to provide for is now less than it was 10 years ago.

What really matters is how you deal with future financial adversity. If you can adjust your expenses down quickly you should be fine.

If you think your ongoing spend will be less than 4% you might consider Audreyh1's approach, which is to withdraw the entire 4%, then save / bank unused funds in a separate account. Or, just set up a separate fund for the those unusual expenses. I do something similar, and it makes it a bit easier to monitor the ongoing spend rate.

 

[audreyh1]

I'm looking at what I withdrew each year compared to the starting balance of that year. I know that's not the basis of SWR concept, now that you mention it. But the portfolio growth has been positive YoY over the past decade despite withdrawals.

I'm sort of thinking that with a 1.7% WR over the past 8 years, it's fairly negligible and perhaps my 'real' start/withdrawal balance began two years ago when I started withdrawing serious percentages.

Since you are actually using % of remaining portfolio method, and not even taking 4% of that most of the time, I wouldn’t worry about it.

For a 50/50 portfolio, you can even take out 4.35% before your income starts to shrink on average as you age.

And 6% won’t run out of money either. On average, however, your real income will gradually shrink as you age.

 

[W2R]

Is my thinking out in the weeds or does this make sense? Can you/should you average your withdrawals over time?

Marko, we all have years when we need more than we did in previous years. My withdrawal rates since retirement have been remarkably similar to what you have described.

I use the "percent of the prior Dec 31st portfolio balance" method. My withdrawal rates in retirement so far have been:

2010: 2.61%
2011: 1.98%
2012: 2.12%
2013: 2.40%
2014: 1.70%
2015: 8.64% (1.72% + 6.92% for extra expenses related to Dream Home purchase)
2016: 1.75%
2017: 1.58%

Believe me, 2015 was a little scary! But then I figured out how much money I had NOT spent, that I could have, had I spent a reasonable 3.5% every year. That more than covered my 2015 spending.

I regarded it as a loan from me, to me. Everything has worked out fine, and my portfolio balance now is more than it was at its highest before buying the house.

Everything is fine. I wouldn't worry if I were you, but also I would try to keep the WR down below 3% for a few years just to be sure.

I think that high spending years like this are easier to absorb during a massive bull market like we have been enjoying, than they would have been back in 2008-2009. So, I would have been more concerned if market conditions were not as they are. In a sense, one might intend to spend more during good years, if one intends to tighten the belt during market crashes.

 

[audreyh1]

I would but I view the past two years as a blip more than a regular thing. My SWR per FireCalc and others is ~4%.

I was more interested if you essentially 'bank' the amount of the 4% that you don't use each year.

Except for the past 2 years (out of my 14 years RE'd) we've barely touched the portfolio (~1.7% per year) and some years not at all; I don't really consider us having started our SWR.

So maybe what I do need to do is figure out WHEN to hit start for a true read on a 'SWR plus inflation' plan.

Well you modeled the initial portfolio value plus inflation method to choose an SWR, but that’s not what you are actually using. I would encourage you to model the %remaining portfolio method. You no longer have to worry about failures in terms of running out of money, but you do have to be aware that if you stick to not exceeding your selected withdrawal rate, you could have some years of significantly reduced real income if you hit a very bad sequence of events.

 

[audreyh1]

Whatever you have done in the past makes absolutely no difference going forward.

Examine your assets. Determine what rate of withdrawal you would like. Analyze to see if your assets can support that withdrawal rate going forward. If not, adjust accordingly. Repeat the process periodically.

Average doesn't matter. People drown in rivers that average 2" in depth.

Joeea is absolutely correct in that what you did in the past makes no difference at all if you look at where your assets are now and choose what is safe enough going forward.

 

[Willers]

We started with a SWR model and have abandoned it to a simple decision each year about what WR seems reasonable based on a % of YE assets. We look at the budget and see what we want to spend and make sure it makes sense. This year it was 3.25% and we're only on pace to spend 2.5%.

We don't bank assets as Audrey does, but we have drifted our stock allocation down from 55% to 50% the last two years so I guess you could say that we are actually indirectly banking it.

We have WR spikes in our plan for car replacements so I don't think an occasional spike matters as long as, mentioned above, the spike doesn't turn into a recurring expense. It doesn't appear that yours will.

 

[marko]

Well you modeled the initial portfolio value plus inflation method to choose an SWR, but that’s not what you are actually using. I would encourage you to model the %remaining portfolio method. You no longer have to worry about failures in terms of running out of money, but you do have to be aware that if you stick to not exceeding your selected withdrawal rate, you could have some years of significantly reduced real income if you hit a very bad sequence of events.

Thanks.
[Edit] Got it. Crap! After 10+ years coming here every day I just figured something new out! What else is out there that I think I know but don't?

Running the "% remaining port" I got a 4.7% rate. I assume I'd be 'forced' to stick to 4.7% for the next 35 years, yes? I used "100%" but didn't see any difference regardless of what I put in.

Following the "95% Rule," from Work Less, Live More, each subsequent annual withdrawal will be the greater of 100% of your previous year's withdrawal, or 4.7% of your current portfolio, with no adjustment for inflation (unlike the normal FIRECalc behavior, which uses your starting portfolio, and makes adjustments for inflation).

Regardless, it does sound like if I under spent for a few years in the future, that unspent % would be available as extra income; same as having set it aside.

 

[audreyh1]

Thanks. I don't understand the underlined above. Do you mean that if I stick to a 5% WR there might be some years that I would need 6% and--technically-- not have it available to me?

Running the "% remaining port" I got a 4.7% rate. I assume I'd be 'forced' to stick to 4.7% for the next 35 years, yes?

Following the "95% Rule," from Work Less, Live More, each subsequent annual withdrawal will be the greater of 100% of your previous year's withdrawal, or 4.7% of your current portfolio, with no adjustment for inflation (unlike the normal FIRECalc behavior, which uses your starting portfolio, and makes adjustments for inflation). Although the calculations are based on unadjusted withdrawals, the charted withdrawals are shown using 2018 dollars.

The idea with % remaining portfolio method, is that you stick to the same % withdrawal every year. The Clyatt method modified this just slightly. You might want to turn that off at first.

Running the % remaining portfolio method in Firecalc does not "pick" a withdrawal rate for you. It tells you how a given withdrawal rate would have performed historically. Big difference. You are selecting the withdrawal rate yourself by inputting your portfolio and withdrawal amount.

What I am saying in the underlined part, is that say you go through a bad sequence of returns - inflation goes up, you go through a bear market or two - your portfolio shrinks in real terms. Guess what, so does your real income aka spending power (including in the Clyatt case). So - yes it's your latter interpretation. It's important to realize how much it could shrink under the worst case scenarios - as unlikely as they are. Unless you take more out than planned even though the portfolio is way down - but in that case you are not following the models anymore.

Because portfolio values can vary widely every year, so will income based on % of the portfolio each year.

 

[audreyh1]

Firecalc does not pick a withdrawal rate for anybody. It lets you model a withdrawal rate against historical data for given withdrawal/spending methods and given portfolio and several additional caveats.

When you give Firecalc a spending and portfolio value, you are picking the withdrawal rate. It just reports what that calculated withdrawal rate is, and give you several statistics about portfolio survival for a give time period and asset allocation.

You actually have to run it several times to chose a withdrawal rate that you think you can live with after reviewing the various results.

 

[TwoByFour]

Joeea is absolutely correct in that what you did in the past makes no difference at all if you look at where your assets are now and choose what is safe enough going forward.

I have wondered about that model in the past, in which you don't use a single WR that was calculated at year-0 of your retirement, but instead run a new SWR model every year based on your current state (age, assets, AA). It is an interesting idea and might work if your remaining sequence of years are statistically independent of the previous year, but they aren't. There is a positive correlation between any two years' returns. Investors know this as "momentum" - a rising market tends to continue to rise.

That same effect exists for any RIP run and this has been discussed on Bogleheads. People tend to retire during an up market, since that is when they have reached their magic number. But when a RIP model is run, it is started randomly without any bias about how the sequence starts - equal weight is given to starting on an up year or a down year. But that assumption is actually incorrect - people tend to start their retirement on an up market.

I don't know how big of an effect the above is, but it is there. It is probably not a reason to not run a WR calculation every year, just take the results with a grain of salt.

 

[marko]

The idea with % remaining portfolio method, is that you stick to the same % withdrawal every year. The Clyatt method modified this just slightly. You might want to turn that off at first.

Running the % remaining portfolio method in Firecalc does not "pick" a withdrawal rate for you. It tells you how a given withdrawal rate would have performed historically. Big difference. You are selecting the withdrawal rate yourself by inputting your portfolio and withdrawal amount.

Because portfolio values can vary widely every year, so will income based on % of the portfolio each year.

[Edit]
Thanks Audreyh1 and others!

We got a bit off track on withdrawal methodologies -- and I learned something new--but I think my question about 'banking' unspent percentages is now clear to me.

I do think I'll move to a % of portfolio method; as you point out that's what I've been doing unconsciously anyway.

I'm really not that dense...I just act that way sometimes!

 

[NW-Bound]

As long as they are actually exception expenses, and not all lifestyle creep...

For a few years, I pondered the costs of having two homes, and questioned my ability to carry both. If these "one-time" expenses kept repeating, then they would not be truly non-recurring. Would not want to fool myself.

These repairs and upgrades subsided the last 2 years, and my WR is now around 2.6% for the trailing 12 months. Would have been 2.3%, if it were not for an expense that would go away in next year.

PS. If the expenses kept growing, I would have to seriously think about early SS.

 

[Dtail]

Thanks.
[Edit] Got it. Crap! After 10+ years coming here every day I just figured something new out! What else is out there that I think I know but don't?

Running the "% remaining port" I got a 4.7% rate. I assume I'd be 'forced' to stick to 4.7% for the next 35 years, yes? I used "100%" but didn't see any difference regardless of what I put in.

Following the "95% Rule," from Work Less, Live More, each subsequent annual withdrawal will be the greater of 100% of your previous year's withdrawal, or 4.7% of your current portfolio, with no adjustment for inflation (unlike the normal FIRECalc behavior, which uses your starting portfolio, and makes adjustments for inflation).

Regardless, it does sound like if I under spent for a few years in the future, that unspent % would be available as extra income; same as having set it aside.

It should be noted that "4.7%" just represents your spending divided by your portfolio balance. It is not your true spending WR% each year, once you include other items such as SS/Pensions/one time expenses, etc.